Abductive reasoning is a multifaceted process, encompassing various perspectives, models, and logics, which can make it sometimes quite intricate to understand. Lorenzo Magnani’s work has been invaluable in providing important distinctions that shed light on relevant aspects of abductive reasoning that I analytically discuss. This chapter mainly investigates both the Peircean account of abduction and the methodology of inference to the best explanation, elucidating the role of cost-benefit analysis within Peirce’s views on the economy of research as well the role of multicriteria analysis within the framework of inference to the best explanation. Then, I introduce a new distinction between strong and weak hypotheses and analysed them in their roles in the context of abduction and inference to the best explanation, respectively. A weak hypothesis is considered justified in the absence of complete evidence against it (given a specific knowledge base), transforming into a conjecture upon a successful abduction. A strong hypothesis is justified by its greater probability compared to the hypothesis of its negated content (given the same knowledge base), aligning with the conservatism criterion in inference to the best explanation. The discussion concludes with the potential and challenges of achieving genuine creative abductions through generative AI systems.